Sixty-four percent of Harvard Faculty of Arts and Sciences respondents now say AI has had a “somewhat negative” or “very negative” effect on their courses, up from 42% just one year earlier, according to The Harvard Crimson’s annual FAS faculty survey published September 14, 2026. That 22-point shift in twelve months is the sharpest single-year move the survey has recorded. Readers curious about AI-powered websites that boost productivity may find the contrast with faculty skepticism striking.
The survey drew 463 responses from more than 1,400 faculty contacted across FAS’s academic divisions and the School of Engineering and Applied Sciences, and the findings carry weight: nearly nine in ten faculty members reported receiving student work they knew or believed was produced using AI.
Explicit AI policies are now nearly universal across FAS. Only 3.5% of faculty reported having no explicit policy, down from 10% the prior year. Full bans rose to 25% of faculty, up from 20%, while fully permissive policies dropped to just 4%, down from 8%.
Sixty-four percent of faculty respondents said AI had a ‘somewhat negative’ or ‘very negative’ effect on their courses this year, a sharp uptick from last year’s survey, when 42 percent said the same.
The Harvard Crimson, September 14, 2026
Detection remains an unresolved problem. Only 64% of faculty said they felt confident distinguishing AI-generated work from student-created work, and just 12% referred a case of unauthorized AI use to the honor council in the past year.
The Discipline Divide Is Real
Social science faculty are the most permissive group; humanities professors are pushing back hardest.
Social science faculty lead in permissiveness, with 84% allowing at least some AI use in their courses. Science and engineering faculty follow at 79%. Arts and humanities faculty sit at 63%.
A separate Crimson analysis of roughly 600 fall 2026 courses sharpens that picture. More than 50% of STEM classes allowed some AI use, compared with 35% of social science courses and only 27% of humanities courses, suggesting that course-level rules in some divisions run stricter than faculty attitudes alone imply.
Harvard College Dean David Deming has urged faculty to accept or even promote AI use in writing- and project-heavy courses. He has simultaneously called for “AI-proof” assessments, such as more in-person exams and assignments designed to resist AI shortcuts. President Alan Garber praised AI as a research tool in a first-day-of-classes address, putting institutional leadership visibly at odds with where faculty sentiment is moving.
Courses Are Being Rebuilt Around the Problem
Professors are not waiting for a detection breakthrough; they are redesigning how work gets done.
Faculty responses reported by the Crimson include requiring drafts and version histories of written work and increasing in-person and oral exams. Process-focused assignments that are harder to complete via AI are replacing traditional take-home writing in a growing number of courses. Broader concerns about chatbot behavior are also gaining attention, as explored in reporting on how Chatbots Stopped Answering questions in expected ways.
The concern has moved from abstract to personal. In 2023, roughly 47% of Harvard faculty predicted AI would have a negative impact on higher education overall; today, 64% say it already has in their own courses. For students, that shift is showing up as new course requirements: expect oral defenses, mandatory draft submissions, and timed in-person assessments to become standard features of the FAS experience.
